<h1 id="california-congressional-districts">2010 California
Congressional Districts</h1>
<h2 id="redistricting-requirements">Redistricting requirements</h2>
<p>In California, according to the <a
href="https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?lawCode=CONS&amp;division=&amp;title=&amp;part=&amp;chapter=&amp;article=XXI">California
Constitution Article XXI</a>, districts must:</p>
<ol type="1">
<li>be contiguous</li>
<li>have equal populations</li>
<li>be geographically compact</li>
<li>preserve city, county, neighborhood, and community of interest
boundaries as much as possible</li>
<li>not favor or discriminate against incumbents, candidates, or
parties</li>
<li>comply with the federal Voting Rights Act</li>
</ol>
<h3 id="algorithmic-constraints">Algorithmic Constraints</h3>
<p>We enforce a maximum population deviation of 0.5%. We use a
pseudo-county constraint to limit the county and municipality splits. We
add VRA constraints encouraging Hispanic VAP and Asian VAP majorities in
districts.</p>
<h2 id="data-sources">Data Sources</h2>
<p>Data for California comes from the ALARM Project’s <a
href="https://alarm-redist.github.io/posts/2021-08-10-census-2020/">2020
Redistricting Data Files</a>. Data for the 2010 California enacted
congressional map comes from <a
href="https://redistricting.lls.edu/state/california/?cycle=2010&amp;level=Congress&amp;startdate=2012-01-17">All
About Redistricting</a>.</p>
<h2 id="pre-processing-notes">Pre-processing Notes</h2>
<p>Islands were connected to their nearest point within county on the
mainland.</p>
<h2 id="simulation-notes">Simulation Notes</h2>
<p>We sample 25,000 districting plans in each cluster across 2
independent runs of the SMC algorithm. We next sample 50,000 districting
plans for California across 2 independent runs of the SMC algorithm for
the remainder. We then thin the sample to down to 5,000 plans. To
balance county and municipality splits, we create pseudocounties for use
in the county constraint. These are counties are Alameda County, Contra
Costa County, Fresno County, Kern County, Los Angeles County, Orange
County, Riverside County, Sacramento County, San Bernardino County, San
Diego County, San Francisco County, San Joaquin County, San Mateo
County, Santa Clara County, and Ventura County, which are larger than a
congressional district in population. A small population tempering value
was used for each cluster to avoid losing diversity at the final step
based on initial runs.</p>
<h3 id="clustering-procedure">1. Clustering Procedure</h3>
<p>First, we run partial SMC in two pieces: the south and the Bay Area.
The counties in each cluster are: - South: Los Angeles, San Bernardino,
Orange, Riverside, San Diego, and Imperial - Bay: Alameda, Contra Costa,
Fresno, Kings, Madera, Madera, Merced, Monterey, Sacramento, San Benito,
San Francisco, San Joaquin, San Mateo, Santa Clara, Santa Cruz, Solano,
Stanislaus, Tulare, and Yolo</p>
<p>We sample in each of these regions with a population deviation of
0.5%. We sample 28 districts in the southern region and 14 districts in
the Bay Area. Because each cluster will have leftover population, we
apply an additional constraint that incentivizes leaving any unassigned
areas on the edge of these clusters to avoid discontiguities. For each
cluster, we add VRA constraints encouraging Hispanic VAP and Asian VAP
concentrations in districts, in line with the enacted plan.</p>
<h3 id="combination-procedure">2. Combination Procedure</h3>
<p>Then, these partial map simulations are combined to run statewide
simulations. We sample 11 districts in the remainder.</p>
<h2 id="contents">Contents</h2>
<ul>
<li><code>CA_cd_2010_stats.csv</code> contains summary statistics on the
sampled redistricting plans</li>
<li><code>CA_cd_2010_plans.rds</code> is a compressed
<code>redist_plans</code> object, which contains the matrix of
precinct/block assignments and may be used for further analysis.</li>
<li><code>CA_cd_2010_map.rds</code> is a compressed
<code>redist_map</code> object, which contains the precinct/block
shapefile and demographic data.</li>
</ul>
<p>Both the <code>redist_plans</code> and <code>redist_map</code> object
are intended to be used with the <a
href="https://alarm-redist.github.io/redist/">redist package</a>.</p>
<h3 id="codebook-for-summary-statistics">Codebook for summary
statistics</h3>
<ul>
<li><code>draw</code>: unique identifier for each sample. Non-numeric
draw names are real-world plans, e.g., <code>cd_2010</code> for an
enacted 2010 plan.</li>
<li><code>district</code>: a district identifier. District numbers
roughly match those in the enacted plan, but the correspondence is not
perfect.</li>
<li><code>chain</code>: a number identifying the run of the
redistricting algorithm used to produce this draw. Used for diagnostic
purposes.</li>
<li><code>pop_overlap</code>: a number indicating the fraction of people
in this plan who reside in the same-numbered district in the enacted
plan.</li>
<li><code>total_pop</code>: the total population of each district.</li>
<li><code>total_vap</code>: the total voting-aged population of each
district.</li>
<li><code>pop_*</code>, <code>vap_*</code>: total (voting-aged)
population within racial and ethnic groups for each district. Variable
codes documented <a
href="https://github.com/alarm-redist/census-2020#data-format">here</a>.</li>
<li><code>plan_dev</code>: the maximum population deviation among
districts in the plan. Computed as
<code>max(abs(distr_pop - target_pop)/target_pop)</code>.</li>
<li><code>comp_edge</code>: compactness, as measured by the fraction of
internal edges kept. Higher values indicate more compactness.</li>
<li><code>comp_polsby</code>: compactness, as measured by the
Polsby-Popper score. Higher values indicate more compactness.</li>
<li><code>county_splits</code>: the number of counties which belong to
more than one district.</li>
<li><code>muni_splits</code>: the number of Census Designated Places
which belong to more than one district.</li>
<li><code>*_##_dem_*</code>, <code>*_##_rep_*</code>: vote counts for
statewide Democratic and Republican candidates in a certain election.
More information <a
href="https://github.com/alarm-redist/census-2020#data-format">here</a>.</li>
<li><code>adv_##</code>, <code>arv_##</code>: average vote counts for
statewide Democratic and Republican candidates in a certain year. More
information <a
href="https://github.com/alarm-redist/census-2020#data-format">here</a>.</li>
<li><code>ndv</code>, <code>nrv</code>: averages of the
<code>adv_##</code> and <code>arv_##</code> variables across all
available elections.</li>
<li><code>ndshare</code>: normal Democratic share, computed as
<code>ndv / (ndv + nrv)</code></li>
<li><code>e_dvs</code>: average Democratic vote share, computed as the
average of the Democratic vote share when first scored under each
statewide election.</li>
<li><code>pr_dem</code>: probability seat is represented by a Democrat;
calculated as the fraction of statewide elections under which the
district had a majority Democratic share.</li>
<li><code>e_dem</code>: expected number of Democratic seats for the
plan; equivalent to summing the <code>pr_dem</code> values across
districts</li>
<li><code>pbias</code>: partisan bias at 50% vote share, averaged across
all available elections. Positive values indicate Republican bias.</li>
<li><code>egap</code>: the efficiency gap, averaged across all available
elections. Positive values indicate Republican bias.</li>
</ul>
